pedma

after the roller coasters i went through last year, im noticing in myself a decreased appetite for risk
im still up 70%+ ytd, not that i missed out, but couldve been better
(probably funny running at 40% vol, having 90%+ of my net worth in my systems in crypto, saying my risk appetite is down)
but i was looking at decent opps months before the move like pump, lit, hype (i did buy some but lower size than i wanted), had a good thesis for them and couldve gotten size, but didn't.
didn't because I kinda hid behind the mantra "let my systems handle it" kinda thing
maybe that's a good thing, maybe

im a systems guy. but waking up to this 3 days in a row is kinda a vibe ya know. the thing that attracted us to crypto in the first place. lets see if it sticks form a while. nothing changed, its all a grift anyways. but ya know, nostalgy feels nice sometimes.

a lot to do
GM

one of the reasons for the increase of output last week was me relaxing the controls over the agents
I was much more willing to let it work on a full task from beginning-end without my intervention, allowing me to target more stuff, but more superficially
i found a decent chunk of bugs that were introduced by some of these changes
that could be a skill issue on my part in setting strict rules for these agents, but also there's other risks when you let them roam free
for example when there's a nerf of a model, you're kinda basing your assumptions off past behavior that is no longer there
and i

really good day yesterday
what im more excited about though is that my leverage was cut in half from almost 5x
i was getting a bit concerned about the different risks that could blow on my face, although being "mkt neutral"
risks arent ignored, noted and will look into it

waking up to fix undocumented api changes at 2 am
gm

im up for the day
although

some time off from the beta mines

always reassuring when opus just answers:
"I made it up"

i keep strict versioning of the engines that drive target exposures for my portfolio
obviously I improve it over time, but i dont want to be working on the engine as the live one is issuing exposures right
also i want the new engine to be tested over a few weeks/months depending on the severity of the rework
but once i want to migrate it to live, i have this template , all i have to do is call the agent and issue the migration
a) old version is archived (for audit purposes)
b) new working version is created
c) full migration to live
its pretty accurate because overtime this template grew bigge

if you document all changes to your codebase , the next time you work on something you already did in the past, the agent can get context from it (if its appropriately tagged and categorized) and warn you about potential pitfalls that we already went through. useful because it makes it harder to run into similar problems.